Page 47 - SRPSKO DRUŠTVO ISTRAŽIVAČA RAKA

SDIRSACR Oncology Insights

of published review articles in the field of medical sciences has dramatically increased over the past ten years and has
rapidly grown in parallel with knowledge in biology and medicine (1, 2).

Review article types
Pre-writing process of a good-to-great-quality review starts with targeting an issue through defining a scientific
(research) question. The issue an author wants to elucidate and resolve through the review should be coherent with
the current problems of the research field, which should take a central place in the manuscript. In practice, it starts with
an idea, inspiration, creativity, and eagerness of the author (usually an expert in the field) to share, acquire, systematize,
reshape knowledge, and spice it up with potential solutions or directions on how to overcome the problems. The story
behind the brilliant idea should be elaborated at the beginning, and ways to overcome a scientific problem should
be mentioned throughout the article. The next step usually begins with the choice of the most suitable review type.
According to readers’/researchers’ information requirements, several types of reviews have evolved. In general, the
three review article types are the most frequent in the research community: traditional or narrative review, systematic
review, and meta-analysis (3).

Narrative review is a type of article where knowledge in the field is harmonized into a coherent unit. A traditional review
article should be well-framed and shaped in a meaningful story based on up-to-date article information integration.
Hierarchically based on evidence importance and impact, it is beneath systematic review and meta-analysis. However,
this does not mean that it is less worthy and important (4). It may be a good basis for the researchers to systematize
their previous findings and compare with recent findings in the field, and to gain new knowledge for planning future
research directions. There are some variations on a review article theme, such as critical review, integrative review,
and many others (4).
The core of the systematic review is evidence synthesis. Evidence synthesis is a scientific approach that systematically
combines findings from previous studies selected according to strict criteria. The major goal of the systematic review
is to draw broader, panoramic conclusions based on carefully summarized data. The key element is defining precise
exclusion and inclusion criteria for the selection of the studies to avoid bias and misinterpretations (5).

Meta-analysis is an approach that statistically analyzes and combines results from more than one study. It is usually
used to estimate the precision and effectiveness of the study results, especially when a small sample size is analyzed in
an individual study. It is usually interpreted and visualized by a forest plot diagram containing parameters, such as the
weight of each included study and its confidence interval (6).

Umbrella review is frequently described as a review of other reviews. It is usually written in the form of an overview
and synthesis of meta-analyses, systematic reviews, and clinical trials, and based on this, it has the highest ranking
not only in the “review field”, but also in the medical field, serving as the summary of the most relevant and the latest
evidence that may be used in clinical practice for decision-making (7). A review article should be distinguished from
“position statement” (which is more in the form of a guideline), though both use similar approaches in the writing
process.

Key steps in writing an effective literature review article
When the topic is clarified and the major research question is defined, the literature should be carefully surveyed and
selected. The literature selection process starts with bibliographic database mining. Setting appropriate keywords and
properly using Medical Subject Headings (MeSH) is crucial. Choosing inadequate keywords can significantly prolong
survey time and deviate from the topic. The most popular electronic bibliographic databases are Medical Literature
Analysis and Retrieval System Online (MEDLINE), Scopus, Excerpta Medica database (EMBASE), Cochrane Library,
Cochrane Register of Trials, Web of Science, etc. (8). PubMed, on the other hand, is a great search engine with links to
published articles, while PubMedCentral (PMC) is a database of deposited articles (9). It is always a better choice to
explore and cross through more than one database when writing a systematic review or meta-analysis. Selection of
the literature is based on the authors’ experience and preferences. Some criteria may involve publication date range
(for example, last ten or five years); type of article (for example, to filter only systematic reviews and/or meta-analysis;
clinical studies, etc.); type of journal (with specific limited scope, or multidisciplinary); impact of the journal (which is
not so relevant criteria, because some great quality articles may be published in lower impact journal and vice versa).
The title of the review is the first thing that affects and attracts the reader. For the first draft of the title, keywords may
be used. Later, when the manuscript gains its shape and matures, the title can be upgraded into an informative, trendy
title, clearly reflecting the topic, focus, issues, and novelty.
The next step should be setting the structure of the review. Authors should determine how to describe and summarize
what has been found, and then start reading and collecting literature. After the critical assessment of the literature,

32